Merging several variables together in perl for RNA alignment - perl

Question

My aligned RNA sequence is
gccuucgggc
gacuucgguc
ggcuucggcc
For which I have done the following coding
open(RNAalign, $ARGV[0]) || "Can't open $ARGV[0]: $!\n";
while ($line = <RNAalign>) {
chomp ($line);
push (#line, $line);
}
#covariences=();
foreach $i (#line) {
foreach $j (#line) {
unless ($i eq $j) {
#search1=split("",$i);
#search2=split("",$j);
$k=0;
while($k<scalar(#search1)) {
if (#search1[$k] ne #search2[$k]) {
$string="";
$string="$k: #search1[$k] #search2[$k]\n";
push (#covariences, $string);
}
$k++;
}
}
}
}
This gives me, when printing:
1: c a
8: g u
1: c g
8: g c
1: a c
8: u g
1: a g
8: u c
1: g c
8: c g
1: g a
8: c u
What I want to do is to merge all the similar position while at the same time keeping any different character that they might have. As following (the characters does not have to be in that exact order):
1: c a g
8: g u c

Answer 1

You can create a hash of hashes (HoH) where the keys are the positions and the values are references to hashes whose keys are the characters at each position. A partial structure of your dataset would be the following:
'8' => {
'c' => 1,
'u' => 1,
'g' => 1
},
'1' => {
'c' => 1,
'a' => 1,
'g' => 1
},
'4' => {
'u' => 1
},
Here's code that produces this HoH structure:
use strict;
use warnings;
my ( %hash, $stringNum );
while (<DATA>) {
chomp;
my $i = 0;
$stringNum++;
$hash{ $i++ }{ lc $_ } = 1 for split //;
}
for my $position ( sort { $a <=> $b } keys %hash ) {
if ( keys %{ $hash{$position} } == $stringNum ) {
my #chars = keys %{ $hash{$position} };
print "$position: #chars\n";
}
}
__DATA__
gccuucgggc
gacuucgguc
ggcuucggcc
Output:
1: c a g
8: c u g
In the while loop, the number of strings is counted, and each string is split into its characters to create a HoH. In the for loop, if the number of keys (e.g., "c") equals the number of total strings, each string varies at that position, so the position and those are printed as an instance of covariance.
Hope this helps!

Answer 2

use strict;
use warnings;
use Data::Dumper;
my $s = "gccuucgggc
gacuucgguc
ggcuucggcc";
print "$s\n\n";
my $data = [];
my #lines = split(/\n/,$s);
chomp(#lines);
my $row=0;
my $col=0;
foreach my $line (#lines){
my #chars = split("",$line);
$col = 0;
foreach my $char (#chars){
$data->[$row]->[$col] = $char;
$col++;
}
$row++;
}
#print Dumper($data,$col,$row);
for(my $i=$col-1;$i>=0;$i--){
my $no_diff = 0;
my $result='';my $prev='';
#print "i: $i\n";
for(my $j=$row-1;$j>=0;$j--){
#print Dumper([$i,$j,$prev,$result,$data->[$j]->[$i]]);
if ($prev eq $data->[$j]->[$i]){
$no_diff++;
}
$result .= $data->[$j]->[$i];
$prev = $data->[$j]->[$i];
}
print "$i: $result\n" if !$no_diff;
}